MNV strain MNV-1 (GV/MNV1/2002/USA), plaque-isolate CW3 (GenBank accession no. EF014462)  was used at passage 6. MNV strain MNV-4 (GV/MNV4/2005/US; GenBank accession no. DQ223043), a field isolate , was used at passage 3. MNV strain CR6 (GV/CR6/2005/US; GenBank accession no. EU004676) was plaque purified three times as previously described for MNV-1  and used at passage 3. Virus lysates were concentrated by ultracentrifugation and resuspended in phosphate-buffered saline (PBS) as described previously . Mock-infected lysates were prepared in a similar manner and used as controls.
Swiss Webster (outbred strain) infection and sample collection
Swiss Webster infection studies were performed as two separate experiments. For both experiments, 7-week-old male outbred Swiss Webster mice were purchased from Charles River Laboratories. All mice were seronegative for MNV at the onset of the experiments. Mice for each experiment were purchased in separate batches and are thus described separately. Mice were housed at the University of Michigan animal facilities. Animal studies were performed in accordance with local and federal guidelines as outlined in the ‘Guide for the Care and Use of Laboratory Animals’ of the National Institutes of Health. The protocol was approved by the University of Michigan Committee on Use and Care of Animals (UCUCA no.: 09710). The first experiment used nine replicate cages, with three mice cohoused per cage and per treatment group. Mice were infected orally with 1 × 107 plaque forming units (pfu) of MNV-1 (n = 27) in a volume of 25 μl PBS, or mock lysate (n = 27) as a control. Tissue samples from the distal ileum, cecum, and colon were collected on days 7, 28, and 56 post infection. Intestinal contents were removed, and half of the cecum and 1 cm pieces of ileum and colon were snap frozen and stored at −80°C.
In the second experiment a total of six replicate cages with three mice cohoused per cage were used per treatment group. Mice were orally infected with 3 × 106 pfu of either MNV-1 (n = 18) or MNV-4 (n = 18) in a volume of 30 μl PBS or mock lysate (n = 18) as a control. Tissue samples from the distal ileum, cecum and colon were collected on days 1 and 3 post infection as described above. Infection was verified by measuring viral shedding in fecal pellets collected at the time of tissue harvest by Charles River Laboratories Research Animal Diagnostic Services (Wilmington, MA, USA) using quantitative reverse transcription real-time polymerase chain reaction (qRT-PCR) as described previously .
DNA extraction (Swiss Webster)
DNA was extracted from murine intestinal tissues using the Roche MagNA Pure Compact system (Roche Diagnostics GmbH, Mannheim, Germany). Tissue samples were suspended in a mixture of Roche Bacterial Lysis Buffer (Roche Diagnostics GmbH, Mannheim, Germany) and sterile PBS (pH 7.4, Invitrogen, Carlsbad, CA, USA). Tissue was disrupted by bead-beating for 1 minute, followed by treatment with 50 μl of proteinase K for 10 minutes at 65°C. Samples were further disrupted by 1 minute of additional bead beating, before heat inactivation at 95°C for 10 minutes. Additional sample processing was performed on the MagNA pure compact system according to the Roche MagNA Pure Nucleic Acid Isolation Kit I protocol (Roche Diagnostics GmbH, Mannheim, Germany). Extracted DNA was quantified on a NanoDrop 1000 spectrophotometer (NanoDrop, Wilmington, DE, USA) and stored at −20°C.
Pyrosequencing (Swiss Webster)
Tissue-derived DNA samples were submitted for 16S rRNA gene amplification and pyrosequencing in two separate batches at the Human Genome Sequencing Center at Baylor University College of Medicine in Houston, TX, USA and the University of Michigan DNA Sequencing Core in Ann Arbor, MI, USA. The V3 to V5 region of the 16S rRNA gene was amplified following the Broad HMP protocol (HMP MDG Default Protocol v4.2), which can be found at: http://www.hmpdacc.org/doc/16S_Sequencing_SOP_4.2.2.pdf.
Amplified PCR products were checked for quality on a 2% agarose gel for visual verification, and each sample was individually quantified using the Quant-It PicoGreen dsDNA kit (Molecular Probes, Eugene, OR, USA). Each sample was diluted to normalize concentrations before pooling. The pooled sample was then checked on a Bioanalyzer 2100 machine (Agilent, Santa Clara, CA, USA) using a DNA1000 lab chip (Agilent) to verify sample purity prior to amplification by emulsion PCR and pyrosequencing.
C57BL/6 (inbred strain) infection and sample collection
In an independent experiment performed in parallel, 7-week-old male C57BL/6 mice (n = 20), five siblings from four sets of identified parents, were used. The mice were bred at Washington University School of Medicine under specific pathogen-free conditions  in accordance with all Federal and University policies. Mice were divided into four treatment groups with one to two mice from each sibling group placed into each treatment group, including untreated control (n = 5), mock infected control (n = 5), MNV-1 infected (n = 5), and CR6 infected (n = 5). All mice were singly housed for the duration of the experiment. All groups except the untreated control mice were given either sterile PBS (25 μl) for mock infection, or orally infected using 3 × 107 pfu of CR6. Feces were collected from each mouse on days 0, 1, 2, 4, 8, 28 and 42 following infection. All mice were seronegative at the initiation of the experiments and as expected only the MNV-infected mice were seropositive at the conclusion of the experiment.
DNA extraction (C57BL/6)
Fresh fecal samples were harvested into sterile screw-top 2 ml tubes containing 500 ml of 0.1 mm zirconia/silica beads (BioSpec, Bartlesville, OK, USA) and stored at −80°C. DNA was isolated by phenol-chloroform extraction and cleaned up using the AMpure (Agencourt, Beckman Coulter, Inc., Brea, CA, USA) system according to the manufacturer’s protocol. The DNA was then diluted to a concentration of 10 to 100 ng/μl.
Each DNA sample was setup in triplicate and pooled at the end of each PCR run to avoid founder effects. The primers used were described previously  and amplified the V1 to V2 region of the 16S rRNA encoding gene, with a universal forward primer and a reverse primer with the addition of a linker and an eight base pair barcode (27F-TCAGAGTTTGATCCTGGCTCAG, 338R-NNNNNNNNCATGCTGCCTCCCGTAGGAGT). The DNA was amplified using 5 Prime Hotmastermix (5Prime Inc., Gaithersburg, MD, USA) with cycling conditions identical to those published previously . The pooled PCR products were run on an agarose gel to confirm the generation of a 300 bp product prior to the submission to the Genome Sequencing Center (GSC) at Washington University, St Louis, USA, for emulsion PCR amplification and 454 Pyrosequencing.
Pyrosequencing data processing and analysis
Analysis of all 454 pyrosequencing data was performed using mothur (version 1.20) . The standard operating procedure instructions for pyrosequencing data processing on the mothur website were followed and can be found at: http://www.mothur.org/wiki/Schloss_SOP.
Pyrosequencing data for both Swiss Webster mouse experiments were analyzed together, while data from C57BL/6 mice were analyzed separately. Both data sets were then processed analogously. First, mothur was used for assigning operational taxonomic units (OTUs), community structure comparisons and classification of 16S rRNA sequences. Classifications were determined by comparing sequences to the Ribosomal Database Project (RDP) (Michigan State University, East Lansing, MI, USA) . Next, sequences were trimmed to remove any ambiguous base calls, those with more than 8 homopolymers, and those with an average quality score below 35 in a window of 50 nucleotides. After trimming, sequences were filtered based on size, and all reads less than 194 nucleotides (Swiss Webster data set) or 219 nucleotides (C57BL/6 data set) were removed. Filtering thresholds were determined based upon alignment with the silva alignment database (http://arb-silva.de/) for each data set. Specifically, during the column alignment, the vertical setting in mothur was set to ‘true’ to remove any column containing gap characters, and the trump setting was set to ‘.’, so any column containing a blank nucleotide was also removed during filtering. Additionally, reads with one or more ambiguous calls were also removed. Only sequences containing the reverse primer were used in the analysis.
OTUs were assigned to sequences based on 97% sequence identity prior to RDP classification. Classified OTUs were used to determine the relative abundance of bacterial phyla in each sample and for statistical comparisons between samples.
Principal coordinates analysis (PCoA) was used to assess community similarity among samples by representing the relative abundance of OTUs in each community using two different analyses. First, a Yue and Clayton-based distance matrix , which measures community structure by incorporating both membership and abundance, was generated. Second, OTU data was analyzed using PCoA via the Jaccard dissimilary index , which measures only membership. These distances were displayed visually in two-dimensional space in PCoA plots. Communities clustered similarly by tissue site, but not by infection status or time, using both Yue and Clayton and Jaccard indices, thus the latter analysis is not shown.
Non-metric multidimensional scaling (NMDS) was used to assess community similarity. NMDS-based ordinations served as a comparator to PCoA to view the data with reduced distortion due to the horseshoe effect seen in PCoA. NMDS values were calculated using the NMDS command in mothur, using all default parameters, and those values were displayed visually in two-dimensional space.
Phylotype analysis was performed using relative abundance values for comparison at both the phylum and family taxonomic levels once relative abundance measurements were averaged for replicate samples in each tissue and treatment group.
DNA sequences are publically available via MG-RAST. Swiss Webster mice are available at: http://metagenomics.anl.gov/linkin.cgi?project=3128. Barcodes for individual Swiss Webster sequences are included in Additional file 1: Table S1. C57BL/6 mice sequences are available at http://metagenomics.anl.gov/linkin.cgi?metagenome=4506745.3. Barcodes for individual C57BL/6 samples are included in Additional file 2: Table S2.
Neither study design measured the bacterial density, so it remains unknown whether MNV infection can raise or lower overall bacterial densities in the intestine.
The relative abundance values of specific bacterial phyla were compared using both the Kruskal-Wallis test of variance and the Mann–Whitney U test. Metastats , a statistical test used to determine differentially abundant features, was used within the mothur bioinformatics program to determine significant differences (at P = 0.05) of specific OTUs in treatment groups based on 3% OTU definitions. All other statistical analyses were performed using GraphPad Prism version 5 (GraphPad Software, San Diego, CA, USA). Measurements of community richness (Chao and ACE) and diversity (Shannon and inverse Simpson) were calculated using mothur, and were based on a 3% OTU definition. ACE richness estimates were based on an OTU with ten or more individuals in it being considered abundant. The number of total OTUs was determined by the sobs calculator in the summary.single command within mothur. It was also based on a 3% OTU definition. These values, and richness and diversity estimates are included in Additional file 1: Table S1 and Additional file 2: Table S2.
When comparing phylotype abundance in either Swiss Webster or C57BL/6 mice data, individual timepoints were combined so comparisons could be made between treatments. This was done because the communities were stable over time, both in Swiss Webster and C57BL/6 mice, and showed little variation throughout the course of the experiment. Specifically, all times were combined in order to compare relative abundance of specific phylotypes between each treatment group. Furthermore, times were also combined for comparison at the OTU level between each treatment group.